Table 2.
Estimated effects of 1998 legislation on poisoning deaths involving paracetamol, for people aged 10 years and over in England and Wales in 1993-2009. Data are mean estimated no of deaths per quarter year, unless stated otherwise
| Estimate of absolute effect (October 1998 to 2009)* | Conservative† estimate of absolute effect (October 1998 to 2009)* | ||||||
|---|---|---|---|---|---|---|---|
| Without legislation‡ |
With legislation‡ |
Difference (95% CI)§ | Without legislation‡ |
With legislation‡ |
Difference (95% CI)§ | ||
| Suicide, open death | |||||||
| Paracetamol | 39 | 22 | −17 (−25 to −9) | 34 | 22 | −12 (−20 to −4) | |
| All drug poisoning (except paracetamol) |
350 | 263 | −87 (−145 to −29) | 321 | 263 | −57 (−115 to 1) | |
| Paracetamol (adjusted)¶ | 28 | 14 | −14 (−23 to −5) | 24 | 14 | −10 (−19 to −2) | |
| Paracetamol (adjusted)** | 20 | 4 | −16 (−25 to −8) | 16 | 4 | −12 (−20 to −3) | |
| Paracetamol compounds | 5 | 6 | 1 (−5 to 7) | 5 | 6 | 1 (−5 to 8) | |
| All drug poisoning | 388 | 285 | −103 (−166 to −40) | 355 | 285 | −70 (−132 to −7) | |
| All causes | 1228 | 1198 | −30 (−199 to 140) | 1252 | 1198 | −54 (−224 to 115) | |
| Suicide, open death, accidental poisoning | |||||||
| Paracetamol | 53 | 31 | −22 (−30 to −14) | 46 | 31 | −15 (−23 to −7) | |
| All drug poisoning (except paracetamol) |
580 | 480 | −100 (−225 to 23) | 513 | 480 | −33 (−158 to 90) | |
| Paracetamol (adjusted)¶ | 41 | 21 | −20 (−32 to −8) | 35 | 21 | −14 (−26 to −2) | |
| Paracetamol compounds | 7 | 9 | 2 (−4 to 8) | 6 | 9 | 3 (−3 to 9) | |
| All drug poisoning | 612 | 493 | −119 (−249 to 11) | 540 | 493 | −47 (−177 to 83) | |
*Using interrupted time series segmented regression analysis,28 where the intervention point is taken as third quarter of 1998.
†Conservative estimate assumes flat trend in post-legislation period had the legislation not occurred.
‡Estimated for the mid-point quarter of October 1998 to 2009.
§Absolute difference of estimated number with legislation and estimated number without, taken at the mid-point of the post-intervention period. Confidence intervals taken from Stata results or calculated according to reference 37.
¶Adjusted by including all deaths from drug poisoning (excluding paracetamol) that received suicide or open verdicts, as a covariate in regression.
**Adjusted by including all suicide and open verdict deaths (except paracetamol) by all methods as a covariate in regression.